Rm62, a DEAD-box RNA helicase, complexes with DSP1 in Drosophila embryos.

Two main classes of proteins, Polycomb group (PcG) and Trithorax group (TrxG), play a key role in the regulation of homeotic genes. These proteins act in multimeric complexes to remodel chromatin. A third class of proteins named Enhancers of Trithorax and Polycomb (ETP) modulates the activity of TrxG and PcG, but their role remains largely unknown. We previously identified an HMGB-like protein, DSP1 (Dorsal Switch Protein 1), which was classified as an ETP. Preliminary studies have revealed that DSP1 is involved in multimeric complexes. Here we identify a DEAD-box RNA helicase, Rm62, as partner of DSP1 in a 250-kDa complex. Coimmunoprecipitation assays performed on embryo extracts indicate that DSP1 and Rm62 are associated in 3- to 12-h embryos. Furthermore, DSP1 and Rm62 colocalize on polytene chromosomes. Consistent with these results, a mutation in Rm62 enhances a null mutation of dsp1 and also mutations of trxG or PcG, suggesting that Rm62 has characteristics of an ETP. We show here for the first time that an RNA helicase is involved in the maintenance of homeotic genes.

DSP1 interacts with bicoid for knirps enhancement.

DSP1 is an HMG-box protein which has been implicated in the regulation of homeotic genes in Drosophila melanogaster. Here we report that DSP1 is also involved in the regulation of the kni gap gene. Analysis of the phenotype of a null mutation of dsp1 (dsp1(1)) reveals that the absence of maternal DSP1 results in A4 segmentation defects that are correlated with a diminution of the kni expression domain. Genetic interaction studies demonstrate that a bcd mutation enhances the A4 defect of dsp1(1). We present in vitro and in vivo evidences for a direct interaction between DSP1 and Bicoid, mediated by the BCD homeodomain and the HMG box of DSP1. Finally, we show by immunoprecipitation of cross-linked chromatin the association of DSP1 with the kni-regulating region and discuss the potential mechanism of DSP1-mediated activation of kni.

DSP1, an HMG-like protein, is involved in the regulation of homeotic genes.

The Drosophila dsp1 gene, which encodes an HMG-like protein, was originally identified in a screen for corepressors of Dorsal. Here we report that loss of dsp1 function causes homeotic transformations resembling those associated with loss of function in the homeotic genes Sex combs reduced (Scr), Ultrabithorax (Ubx), and Abdominal-B. The expression pattern of Scr is altered in dsp1 mutant imaginal discs, indicating that dsp1 is required for normal expression of this gene. Genetic interaction studies reveal that a null allele of dsp1 enhances trithorax-group gene (trx-G) mutations and partially suppresses Polycomb-group gene (Pc-G) mutations. On the contrary, overexpression of dsp1 induces an enhancement of the transformation of wings into halteres and of the extra sex comb phenotype of Pc. In addition, dsp1 male mutants exhibit a mild transformation of A4 into A5. Comparison of the chromatin structure at the Mcp locus in wild-type and dsp1 mutant embryos reveals that the 300-bp DNase I hypersensitive region is absent in a dsp1 mutant context. We propose that DSP1 protein is a chromatin remodeling factor, acting as a trx-G or a Pc-G protein depending on the considered function.

HMG boxes of DSP1 protein interact with the rel homology domain of transcription factors.

Formation of the dorsoventral axis in Drosophila melanogaster is mediated through control of the expression of several genes by the morphogen Dorsal. In the ventral part of the embryo Dorsal activates twist and represses zen amongst others. Recently, several proteins have been shown to assist Dorsal in the repression of zen, one of which is DSP1, a HMG box protein that was isolated as a putative co-repressor of Dorsal. In this report we used a DSP1 null mutant to ascertain in vivo the involvement of DSP1 in Dorsal-mediated repression of zen but not in the activation of twist. We show that Dorsal has the ability to interact with DSP1 in vitro as well as with rat HMG1. Using truncated versions of the proteins we located the domains of interaction as being the HMG boxes for DSP1 and HMG1 and the Rel domain for Dorsal. Finally, studies of the zen DNA binding properties of Dorsal and another related Rel protein (Gambif1 from Anopheles gambiae) revealed that their DNA binding affinities were increased in the presence of DSP1 and HMG1.

DSP1 gene of Drosophila melanogaster encodes an HMG-domain protein that plays multiple roles in development.

DSP1 is an HMG-box containing protein of Drosophila melanogaster which was first identified as a co-repressor of the Dorsal protein. Recently, the analysis of the structure of the gene has led us to propose that DSP1 is the Drosophila equivalent of the ubiquitous vertebrate HMG 1/2 proteins. In the present paper, the patterns of expression of DSP1 protein and RNA in adult flies and during development are reported. In the adults DSP1 protein is located in nurse cells of ovaries and in brain. During eggs development uniform expression of DSP1 protein persists until the end of germband retraction. At later stages, expression is restricted to the ventral nerve chord and brain. Using P-element mutagenesis, we have isolated a mutant deficient in DSP1 functions. Genetic studies of this mutant show that DSP1 protein is essential for the growth and the development of Drosophila. In addition to be a co-repressor of the transcriptional activator Dorsal our results provide compelling evidence that DSP1 is a regulator involved in several pathways necessary for the development of the fly.

The Drosophila DSP1 gene encoding an HMG 1-like protein: genomic organization, evolutionary conservation and expression.

The gene that encodes the dorsal switch protein (DSP1) has been isolated from a Drosophila melanogaster cosmid library. It is organized into seven exons and six introns. The relative position of the introns within the region coding for the high mobility group (HMG) domains are identical to those of vertebrate HMG 1/2 genes. The close similarity between DSP1 and HMG 1/2 genes strongly suggests that these genes derived from a common ancestral gene. DSP1 encodes, at least, two distinct mRNAs that differ in the length of their 5'-untranslated region and coding sequence. Detailed sequence analysis shows that alternative splicing of precursor mRNA gives rise to the two isoform mRNAs found in Drosophila cells.

Large deletions induced in the white gene of Drosophila melanogaster by the antitumoral drug cis-dichlorodiammineplatinum(II): influence of non-homologous recombination.

We have studied two mutants carrying large deletions induced in the white gene of Drosophila by the antitumoral drug cisplatin. The breakpoints of the deletions were located by southern analysis and the sequences of the deletion junctions were determined. Two base-pair repeats are associated with the ends of these deletions; one of the repeats is preserved in the new junction after the deletion. DNA sequences such as A-T rich, alternating purine/pyrimidine tracts, polypurine-polypyrimidine tracts and topoisomerase I and II cleavage sites are found near the junctions. These results suggest that illegitimate recombinational processes are involved in the generation of cisplatin-induced large deletions.

Interaction between cisplatin-modified DNA and the HMG boxes of HMG 1: DNase I footprinting and circular dichroism.

The interactions between the two boxes A and B of HMG 1 and cis-diamminedichloroplatinum(II)-modified DNA containing a single intrastrand cross-link at the d(GpG) site were studied by DNase I footprinting and circular dichroism. The DNAase I cleavage patterns of the HMG box-platinated DNA complexes are identical, the two boxes inhibiting the DNase I cutting over at least 15 and 12 nucleotide residues in the platinated strand and the complementary strand, respectively. As judged by circular dichroism, the two boxes have the same alpha-helical content (56%) and they induce the same conformational changes in the platinated DNA.

Deletions induced in the white and vermilion genes of Drosophila melanogaster by the antitumor drug cis-dichlorodiammineplatinum(II).

This paper describes the analysis of cisplatin induced mutations at the white (w) and vermilion (v) loci located on the X chromosome of Drosophila melanogaster. Twenty-eight w and eight v mutants have been found in a male genetic context and 42 w mutants in a female genetic context. In these latter experiments, genetic analysis showed the presence of multi-locus deficiencies in 18 out of 42 w mutants. Eighteen w and three v intragenic mutations were analyzed at the molecular level. Seventeen w and three v mutants carry deletions within the gene, ranging in size from 4 to 109 base pairs. Sequence analysis of the mutants indicates that most of them were produced by non-homologous recombinational events occurring between short (2-5 bp) sequence repeats on both sides of the deletion, one repeat being retained at the new junction. These results differ largely from those obtained in prokaryotic and other eukaryotic cells.

Detection of minor adducts in cisplatin-modified DNA by transcription footprinting.

Two DNA restriction fragments containing either a d(GC)5 or a d(TTGCTTGATTAGTTGTGTT) insert were subjected to reaction with cis-diamminedichloroplatinum(II) and were then used as templates for RNA synthesis by T7 RNA polymerase. Within the d(GC)5 insert, interstrand cross-links are preferentially formed. Within the second insert, the reactivity order of the potential binding sites is d(ApG) > d(GpC/GpC) = d(GpA) > d(GpTpG). In the presence of cyanide ions, the adducts are much less stable at the d(GpA) sites than at the d(GpCpG) sites, in double-stranded DNA.

Genomic organization and nucleotide sequence of a long mosaic repetitive DNA in the mouse genome.

A long mosaic repetitive sequence (LMRS) was isolated from a mouse liver genome library using a mouse repetitive DNA as a probe. LMRS exhibits the following features: (1) it is almost 15 kb in length; (2) it is partly organized in tandem array and frequently interrupted by other repeated sequences; and (3) it is located predominantly on the A3 band of the mouse X Chromosome (Chr). One fragment of LMRS (B6) shows restriction fragment length polymorphism (RFLP) between different mouse strains, and is thus potentially useful for mapping studies. The nucleotide sequence confirms a mosaic organization of LMRS which includes three repeats in the 5' part, showing similarity with the 5' end of L1Md-A2, and seven long A + T rich segments in the central part of the element. Our findings suggest that this sequence may have arisen from the duplication of an ancestral motif and has expanded by successive waves of amplification and invasion by foreign sequences.

Studies of a novel repetitive sequence family in the genome of mice.

A new middle repetitive sequence is described in the mouse genome. It has been revealed with a recombinant clone isolated from a Mus musculus BamHI gene library constructed in pBR322 and containing an insertion of 1.73 kb. When digests of genomic DNA were subjected to Southern blot hybridization, using the 1.73-kb insert as probe, we obtained a light smear and discrete bands, indicating a dispersion in the mouse genome of this sequence. This 1.73-kb sequence seems to be a part of a greater repetitive sequence at least 6 kb in length. The sizes of the bands hybridizing with the 1.73-kb insert are similar when compared between different laboratory strains but differ remarkably between the two species M. musculus and Mus caroli. We have shown also a great variation in the copy number of the sequence studied between these two species. When rat DNA is probed with the 1.73-kb insert, no hybridization is observed. Subcloning of the 1.73-kb sequence in three fragments has pointed out that the reiteration was not homogeneous along the 1.73-kb sequence. The 1.73-kb clone was sequenced and compared with other interspersed repetitive sequences, previously described in the rodent genome, and no homology was found.

Interacting genetic factors controlling the antibody response against the H-2.2 specificity.

The antibody response against the H-2.2 specificity has been studied in three H-2d strains. B10.D2. DBA/2. and BALB/c. and their hybrids (B10.D2 x DBA/2)F1 and (B10.D2 x BALB/c)F1. The genetic control of the response appears to be complex: The three pure strains are responders, whereas both hybrids when immunized with C3H-HTG are nonresponders. Individual analysis of N3 offspring is compatible with the idea that, in this combination, an Ea-4 incompatibility between donor and immunized strain is necessary for the anti-H-2.2 response to occur. H-2d/H-2k hybrids (B10.BR x B10.D2)F1 or (B10.BR x DBA/2)F1 are responders when immunized with C57BL/10 (H-2b) but not with B10.A(2R) (H-2h), indicating that simultaneously recognized H-2 specificities are necessary for the anti-H-2.2 response.